Structural Equation Modeling: A Multidisciplinary Journal ( IF 6 ) Pub Date : 2022-06-01 , DOI: 10.1080/10705511.2022.2063870 H. Du 1 , P. M. Bentler 1
Abstract
In structural equation modeling, researchers conduct goodness-of-fit tests to evaluate whether the specified model fits the data well. With nonnormal data, the standard goodness-of-fit test statistic T does not follow a chi-square distribution. Comparing T to can fail to control Type I error rates and lead to misleading model selection conclusions. To better evaluate model fit, researchers have proposed various robust test statistics, but none of them consistently control Type I error rates under all examined conditions. To improve model fit statistics for nonnormal data, we propose to use an unbiased distribution free weight matrix estimator () in robust test statistics. Specifically, using normal theory based parameter estimates with we calculate various robust test statistics and robust standard errors. We conducted a simulation study to compare 63 existing robust statistic combinations with the 4 proposed robust statistics with The Satorra–Bentler statistic TSB based on () provided acceptable Type I error rates at .05, or .1 across all conditions (except a few cases with ), regardless of the sample size and the distribution. or typically provided the smallest Anderson-Darling test values, showing the smallest distances between p-values and We use a real data example to compare statistics with and that with
中文翻译:
已有 40 年历史的无偏分布免费估计器可靠地改进了非正态数据的 SEM 统计数据
摘要
在结构方程建模中,研究人员进行拟合优度检验以评估指定模型是否很好地拟合数据。对于非正态数据,标准拟合优度检验统计量T不服从卡方分布。将T与可能无法控制 I 类错误率并导致误导性的模型选择结论。为了更好地评估模型拟合度,研究人员提出了各种可靠的测试统计数据,但没有一个能够在所有检查条件下始终如一地控制 I 类错误率。为了改进非正态数据的模型拟合统计,我们建议使用无偏分布自由权重矩阵估计器() 在稳健的测试统计中。具体来说,使用基于正态理论的参数估计我们计算各种稳健的测试统计量和稳健的标准误差。我们进行了一项模拟研究,将 63 个现有的稳健统计组合与 4 个提出的稳健统计进行比较Satorra–Bentler 统计量T SB基于() 提供了可接受的 I 类错误率.05,或 .1 在所有条件下(除了少数情况),与样本大小和分布无关。或者通常提供最小的 Anderson-Darling 检验值,显示p值和我们使用一个真实的数据示例来比较统计数据那与